Friday, 29 January 2016

Cultural handicaps

The handicap principle has proved to have some merit in the organic realm. Handicaps can be a form of costly signaling. However handicaps don't need to be in the genes. They can be culturally transmitted. This article is about cultural handicaps.

Boys sometimes handicap themselves in order to show off to girls. They ride bicycles with no hands, they get drunk and take risks and they engage in costly displays of public generosity. The handicap principle suggests that public charitable donations will often be made by men.

Ladies also adopt cultural handicaps. They wear high-heeled shoes, crush their waists using corsets, bind their feet, conceal their faces and mutilate themselves.

Cultural handicaps are celebrated in popular culture. Inigo Montoya and The Man in Black both fight with their left hand in The Princess Bride. Luke Skywalker fights blindfold in Star Wars, A New Hope.

Cultural handicaps are also used in the game of go - where one player disables himself in order to have a fair fight with another. Handicapping is common in sports. There's a Wikipedia page about the topic.

Many ostentatious social norm violations could be seen as being handicaps. Real men wear pink is one example.

So far, cultural handicaps do not appear to have received the same level of study as organic handicaps. This is probably because of the immature state of the science of cultural evolution.


My review of Harnessed is up. I covered the book when it came out in 2011, expressing my approval of the symbiology in it and Mark's use of the M word. However, I only just got around to reading the book. At the end of my review I express my approval of the core idea in the book, "harnessing".

Harnessing as a process similar to domestication. You can sometimes harness a wild animal and put it to use without domesticating it. Harnessing is often a prelude to domestication. As an example of the process, Mark claims that dogs have been domesticated, while cats have been harnessed. I see more harnessed dogs than harnessed cats, but Mark observes that saucers of milk count as a type of harness - in the sense in which he is using the term..

Mark applies the concept of harnessing to cultural evolution - claiming that humans have mostly been harnessed rather than domesticated. Many say that humans have been domesticated by cultural institutions. However, Mark is probably right that these are often cases of harnessing - rather than domestication.

I think that "harnessing" is useful terminology. It is also an important idea for students of cultural evolution. Humans are harnessed in many cases, just as Mark claims. Kudos to Mark for figuring this issue out and sharing his conclusions.

Thursday, 28 January 2016

Memetics: too easy to understand

I've long entertained the idea that the unpopularity of memetics in academia might be associated with the fact that it is too easy to understand.

Notoriously, academia is not about learning, understanding or credentialing folks. It is about social networking and affiliating with impressive individuals. Simple topics are shunned by academics for not being impressive enough to interest them.

I may have got the idea that this applies to memetics from Keith Henson. In 2012, Keith wrote me a private email containing the following:

Mostly I find that the academic types want to make the subject so complicated that nobody outside the field can understand it. They can't because the concepts are just too simple. Thus they ignore it and the work gets done by people disconnected from academia. Kind of a people's science.
Maybe I had the theory before talking to Keith - but his email crystallized the idea for me.

I think that this is part of the explanation for some of the dense maths used by cultural evolution enthusiasts. Lumsden and Wilson's 1981 book makes a good example of this. That book is full of mathematical models. However the models do not really serve any useful purpose - instead they represent a sort of showing off. The authors are attempting to make themselves seem more impressive via their mastery of arcane mathematics.

If Sheldon Cooper and Cartman can understand memetics, academics can't be failing to understand it due to the difficulty of the subject matter.

This theory also explains why cultural evolution enthusiasts reinvented a whole bunch of new models - rather than using ones from epidemiology and symbiology. An esoteric form of cultural evolution is harder to learn - and mastery of it is more impressive.

One of the reasons academics criticize memetics is because it is too simple. For example, consider: Meme Theory Oversimplifies Cultural Change. This critical article has it backwards. The authors are over-complicating their theory of cultural change beyond what is necessary - contradicting Occam's razor. Affiliation theories offer a parsimonious explanation of why academics might behave in this way.

Memeticists studied cultural evolution while academics almost all studied gene-meme coevolution. Trying to study gene-meme coevolution without first fleshing out a theory of meme evolution is like trying to fly before you can walk. How did this happen? Prestige / affiliation theories suggest it is because flying is the more difficult and impressive feat.

Prestige / affiliation theories suggest that memes might be more popular in engineering disciplines. Where the rubber hits the road, so to speak. I can point at memetic algorithms, military memetics and meme-based marketing as examples of cases where this might be true.

An affiliation theory also looks promising if you look at the characters that have supported memes over the years. I don't mean to insult my fellow travelers - but meme enthusiasts have been a motley bunch of oddballs and kooks. Memeticists have often not been attractive candidates for other academics to affiliate with. That creates a vicious cycle.

The affiliation theory presented here suggests that the glacial uptake of cultural evolution might be an internally-generated phenomenon. If so, sad times for science and the truth.

The meme turns 40

The Selfish Gene was published 40 years ago. That makes the meme 40 years old.

Matt Ridley has penned an article celebrating 40 years of Selfish Genes. The article is a good one - but it fails to mention the M word.

Memes could turn out to be the most significant contribution made in The Selfish Gene. Dawkins popularized the nascent field of cultural evolution - making it comprehensible to the layman. Five years later four academics published books on the topic, dragging the topic into the academic mainstream, and the rest is history. While the history of cultural evolution long predates The Selfish Gene, retrospectively, 1976 seems like a significant turning point for the field.

Wednesday, 27 January 2016

Branch dynamics

Tree-shaped patterns in nature are frequently found in both the organic and inorganic domains and demand a common explanation. One type of explanation is a functional one. Many trees efficiently drain basins. Electrical discharges can be modeled as draining electrical charge and propagating cracks do something similar with stress. The maximum entropy production principle generalizes this idea so that it can be applied to evolutionary family trees as well as to physical trees.

An alternative approach to modeling natural tree-shaped systems is to enumerate the type of operation that are possible on them, and develop specific theories for each type of operation. This is the approach taken by Darwinism - which breaks the dynamics of family trees into reproduction with selection, mutation and recombination. Darwinism also applies to branch tip dynamics in other tree systems - since branch tips split, mutate, undergo selection and inherit properties from their parent(s). At the very least they inherit their position.

I've argued elsewhere that positional inheritance is ubiquitous and that this means that many inorganic systems exhibit Darwinian dynamics. For example, the tree-shape propagating cracks, electrical discharges and fractal drainage basins are family trees. In these systems, the branching tips of the trees form a population which is subject to copying with variation and selection. As a result conventional Darwinian concepts of fitness and adaptation apply - and these systems act as optimizers in a similar way to what is seen with genetic algorithms.

With positional inheritance, most physical tree-shaped patterns in nature are regarded as being family trees. However the fact that these trees are physical trees (rather than being historical ones) means that their branches might undergo dynamic changes. Because a family tree is historical its branches form a fixed structure which can't be changed once it has formed. However, the branches of other natural tree-shaped patterns (e.g. streams or with electrical discharges) sometimes do change their structure over time. The branches don't just grow at the tips, they wave and break.

What this means is that we need a theory of branch dynamics - in addition to the Darwinian dynamics associated with the population of branch tips.

Basic branch operations that do not change branch topology include:

  • Branch widening;
  • Branch narrowing;
  • Branch lengthening;
  • Branch shortening;
In addition, branches can be destroyed and sometimes entire branches may also split off or join together. Branches can break without becoming detached if the tree is reticulated (as is fairly common in natural systems). Streams illustrate branch widening, narrowing, lengthening and shortening. Streams get wider as more rain falls and they narrow in times of drought. In meandering rivers on plains, streams lengthen. Branch separation is illustrated by crystals and coral. Sometimes a chunk breaks off without being destroyed. It might go on to start a new colony elsewhere. Branch death sometimes follows branch separation. It can also be a consequence of extended branch shortening or narrowing.

Specific theories relating to these branch operations are generally domain specific. However there may be commonalities between the operations in different domains. Modeling such commonalities is part of the study of branch dynamics, though it is beyond the scope of this post.

Tree dynamics are of theoretical interest because trees are common natural objects. It is also noteworthy that it is possible to recover traditional Darwinism from the theory by considering the special case where the tree branches become fixed once they are created and only the branch tips evolve. The principles of tree dynamics are a superset of those in Darwinism.

Monday, 18 January 2016

Diagram of bank mergers

I've long argued for the importance of merging and joining as fundamental operations in both organic and cultural evolution. Cultural evolution is famous for its mergers. Here's a nice example of merging - from economics:

Thanks to Shannon Vyff for drawing my attention to this one.

Sunday, 17 January 2016

Cultural cancer

Cancer is a common phenomenon in the organic realm - and the parallels between organic and cultural evolution lead us to ask: is there such a thing as cultural cancer?

Organic cancers arise when an organisms' components malfunction and reproduce excessively. Often they threaten to kill their owners.

Artifacts do not normally behave in this way. They are usually not made up of self-reproducing entities in the first place - and so the chances of their reproduction getting out of control is low.

However there are some classes of cultural phenomena that resemble cancers. Cancers feature out-of control reproduction of cells that are not supposed to reproduce at all, or are time-bombed to only reproduce a limited number of times. Many cultural items have deliberately limited reproductive potential - enforced by technical barriers to copying, legal barriers to copying. Technical barriers include DRM and dongles. However sometimes these barriers are circumvented.

Seen in this light, counterfeiting might be seen as a type of cultural cancer. Hyperinflation caused by excessive money printing could be regarded as being a form of cultural cancer too. Copying material which is supposed to be protected from copying by copyright law is another example. Reproduction of private documents on the internet after a 'doxing' attack could also be regarded as being a form of cultural cancer.

A related phenomena takes place with memetic hitchhiking. Memetic hitchhiking often involves a viral vector with an attached payload. However, if the payload is not securely attached, it can sometimes become detached - leaving the viral content to reproduce is an unrestrained manner, without distributing the payload. The type of mutation that leads to this separation can be damaging. This phenomenon could also be regarded as a form of cultural cancer.

We should also consider defining cancer to refer to cases where part of an organism threatens the integrity of the whole. In this case, another range of scenarios start to look like cultural cancers. Enron was destroyed by rogue elements from within. The Shakers could be seen as having been extinguished by a type of ideological cancer - as could Heaven's Gate.

One problem with classifying these type of phenomena as being cultural cancers is that they don't occur inside proper organisms. Rather, companies and religious groups could be seen as being symbiotic unions of multiple creatures (both cultural and organic) - rather like a Portugese man o' war. Rather than looking like cancer, disorders involving one part of a composite creature profiting at the expense of the whole just look like ordinary resource squabbles within uneasy alliances.

Saturday, 16 January 2016

Critics of the evolution revolution

I'm a proponent of the idea of a modern evolution revolution. I think that the modern changes required to evolutionary theory amount to a paradigm shift. In particular, the expansion of the domain of Darwinian evolution as described by universal Darwinism represents a revolution of significant size. For example, it requires extensive revisions to the evolution textbooks.

However, I notice that there are naysayers. One is Razib Khan. Here are some of his articles on the topic:

Razib says: "There is no revolution in evolution" ...and: "I would argue there hasn’t been a true revolution in evolutionary biology since Mendelian genetics and classical Darwinism were fused in the 1920s and 1930s".

The problem with Razib as a critic is that he doesn't really address universal Darwinism. I don't think "epigenetics" is much of a revolution either. In fact I would describe "epigenetic inheritance" as being an oxymoron ( e.g. see here and here).

Razib does have some knowledge of cultural evolution. Why doesn't he consider it to be a revolution? The answer isn't clear to me. Razib says: "Genetics began as inferences about the nature and character of inheritance from observed patterns, not by understanding molecular biological mechanisms." The implication seems to be that more types of inherited character are no big deal and that genetics implicitly has them covered. That's a position that I understand and am sympathetic to - but I don't think it means that there's been no revolution.

Is Universal Darwinism really a revolution? Because it has taken so long perhaps it could be described in terms of evolution - rather than revolution. Or maybe it is a revolution that hasn't fully happened yet - and so people don't recognize or understand it. I tend to favor the latter hypothesis. For the most part, the revolution naysayers haven't shown that they understand the post-revolution perspective. Obviously, if you are on the wrong side of the paradigm shift, you don't see any revolution. Let's see the critics write some articles about Darwinian physics or the incorporation of observation selection effects or intelligent design into Darwinism - to prove that they understand the subject - before denigrating the significance of such developments.

Wednesday, 13 January 2016

Neural Darwinism

Theories broadly based on Darwinian evolution apply at various levels in the brain. Long-term memories are copied to preserve them. Learned information is repeatedly rehearsed. Mistakes are repeatedly dwelled on. Promising ideas and plans are copied in short term memory as variations on them are generated and selectively maintained.

Also, various low level processes behave in a Darwinian fashion: most significantly, neurite tips are copied with selective retention, and nerve impulses themselves are copied as they travel down branching axons and up branching dendrites.

I think that it makes reasonable sense to refer to these low-level brain processes using the term 'Neural Darwinism'. There's a bit of a problem though - which is that the term 'neural Darwinism' has been widely used to refer to the particular theories put forth in the late Gerald Edelman's 1987 book titled neural Darwinism.

Gerald Edelman was a pioneer in applying Darwinism to the brain - and I don't want to diminish his contribution too much. Developmental selection is a reasonable idea. Experiential selection is a bit more of a dubiously-named idea. I read neural Darwinism in the 1990s and found it dry, tedious and unconvincing. However, retrospectively, Edelman's concept of neural reentry looks important. It could be the key to understanding how the brain does something functionally similar to back propagation without having a good quality bi-directional signal propagation mechanism.

Overall, though, I think things have moved on a bit since 1987. It is now clearer that there are at least three types of low-level signal copying in the brain: the conventional reproduction of cells - including neural stem cells, the splitting of axon and dentrite growth tips, and the splitting of signals travelling how axons and back-propagating up dendrites. Despite the progress, I think we can still use the term "Neural Darwinism". I would hate "Neural Darwinism" to become a reference to an out-dated and discredited theory: it deserves better than that.

Signals propagation in axons and dentrites has been effectively simulated by synthetic neural network enthusiasts. However few of them bother simulating the other main copying processes in the brain: cell splitting during development and neurite tip splitting. Maybe simulating these Darwinian brain processes would help to build better synthetic neural networks.

For references see my Keeping Darwin in mind essay.

Saturday, 9 January 2016

Jared Diamond argues that religions must be beneficial

Meme enthusiasts like to point out that the common argument that religions must be beneficial to humans - or else they would have gone extinct by now - is a fallacy. I notice that Jared Diamond committed exactly this fallacy in his 2013 book, The World Until Yesterday. Here's what he wrote:

For individuals and for societies, religion often involves a huge investment of time and resources. To mention just a few examples, Mormons are expected to contribute 10% of their income to their church. It's estimated that traditional Hopi Indians devote an average of one out of three days to religious ceremonies, and that one-quarter of the population of traditional Tibet consisted of monks. The fraction of resources in medieval Christian Europe devoted to building and staffing churches and cathedrals, supporting the many orders of monasteries and nunneries, and underwriting crusades must have been large. To borrow a phrase from economists, religion thus incurs "opportunity costs": those investments of time and resources in religion that could have been devoted instead to obviously profitable activities, such as planting more crops, building dams, and feeding larger armies of conquest. If religion didn't bring some big real benefits to offset those opportunity costs, any atheistic society that by chance arose would be likely to outcompete religious societies and take over the world. So why hasn't the world become atheistic, and what are those benefits that religion evidently brings? What are the "functions" of religion?

Essentially, the same argument shows that the common cold must be beneficial, or else those immune to it must have taken over the world. That argument does not prove that the cold virus is beneficial to humans. So, merely noting that something is common doesn't show it is beneficial to humans. Bedbugs are common and they have clear costs - but that doesn't mean that they must have associated benefits that outweigh these costs. The whole argument is just a fallacy.

Friday, 8 January 2016

Charles Goodnight on memes

Charles Goodnight has weighed in again on memes. This time it seems like something of a turnaround.

One of his previous contributions was this critical comment:

I think Dawkin’s concept of the “meme” is particularly telling. As I pointed out culture is intrinsically continuous and no problem for the phenotypic view of evolution; however, for the genic view it is a huge problem. In Dawkin’s view the gene as an object is the center of evolution. Culturally inherited traits cannot be objects from the genic perspective unless they are atomized. The meme is an attempt to force this intrinsically continuous concept into the particulate framework that is essential for the genic view of evolution.

That's much the same thing that many anthropologists say: memetics is an attempt to "atomize" culture.

Of course from the information-theoretic perspective that I prefer, splitting inherited cultural information up into pieces is a trivial operation. You can divide culture into memes in much the same way as you can divide information into bits. That culture can be digitized is seen on the internet - where culture of all kinds is split into streams of 1s and 0s. Of course, splitting things up in order to analyze them is also absolutely standard scientific practice.

It is true that memes sometimes depend on one another and they can't necessarily be analyzed independently. The exact same thing is true of DNA genes. Memes are just like genes in this respect.

Anyway, now Charles seems to have come around to the idea that memes might be a useful concept in cultural evolution after all. The odd thing is that it's the internet meme that he wants to use. Here's what he says:

Interestingly, the internet the term meme, or internet meme, which needs to be distinguished from the Dawkins meme, appears to be directly analogous to the collection of individual concepts that are closely related. Thus, I would suggest that we use internet meme, or just meme, to describe a set of similar individual concepts that are held by individual humans.

Promoting the concept of "internet meme" by shortening it to "meme" was also propsed by Limor Shifman in her book Memes in Digital Culture.

It is hard for me to take these proposals seriously. The idea that Millhouse is not a meme just seems stupid to me. We don't normally define "gene" in terms of how popular a particular sequence is (some comments by G. C. Williams about 'appreciable frequency' notwithstanding) - so if memes are like genes we shouldn't treat them differently in this regard.

I'm mostly OK with people shortening "popular internet meme" to "meme" - and there's certainly a need for a name for popular content. However "meme" is already taken - and it refers to a more significant and important concept.

Tuesday, 5 January 2016

Richard Dawkins explains the real meaning of the word 'meme'

The blurb reads:

Today, the word "meme" is typically used to describe a funny photo with text that gets passed around online. But Richard Dawkins coined the term years before the World Wide Web even existed.

Thanks to Kosmozoan on Reddit for bringing this one to my attention.

Friday, 1 January 2016

An introduction to Darwinian physics

Darwinism's application domain expanded in the 20th century from its traditional area - of the evolution of DNA-based organisms. Despite vociferous objections from some anthropologists and philosophers, it is now widely recognized to cover a wide variety of cultural phenomena - including science, technology, language and religion.

Paralleling the rise of Darwinian cultural evolution, Darwinism was also applied to the development of organisms. First, evolution within the immune system was observed - and then various evolutionary models of other developmental stages were found to be useful. Multi-cellular organisms can be usefully regarded as populations of cells which themselves evolve over the lifetime of the organism they are part of. Darwinian models of psychological evolution were also developed. Skinner's model of learning, for example, was explicitly Darwinian. Gerald Edelman was involved in both the immune and neural breakthroughs.

These expansions of the application domain of Darwinian evolution led to a back-to-basics study of Darwinism and its limits. These studies suggest that Darwinism has applications well beyond biology.

Physics has long laid claim to observation selection effects - though the main students of selection are those studying evolutionary biology. However, it was additionally found that many simple physical systems can be modeled using traditional Darwinian models - based on copying with variation and selection. It turns out that the tree-shaped patterns found in electrical discharges, propagating cracks, and fractal drainage systems are all composed of family trees - representing patterns of descent which are subject to variation and selection. This can be seen in slow motion images of these phenomena.

One of the key concepts in understanding these systems is positional inheritance - or, more generally, spatio-temporal inheritance. When entities split, the offspring often inherit a variety of properties from their parents. One set of properties that are regularly inherited with high fidelity are spatio-temporal coordinates. Other properties are also regularly inherited. Velocity, charge, temperature and chemical composition are other examples.

In practice, many variables correlate with position. A splitting pebble inherits its parent's position, but it also inherits its acidity level, humidity, temperature, and many other parameters from its parent's ecosystem. This is due to the smoothness of nature. Nature's smoothness helps to ensure that fitnesses are inherited - which is one of the requirements for adaptive evolution.

The upshot of all this is that we can use Darwinian models to study simple physical systems, and the concepts of adaptive fit and fitness are applicable. The resulting field of study is known as Darwinian physics, which is considered to be part of Universal Darwinism.

Physicists have long realized that physical laws might cover these types of complex systems - and various models of them have been built. Often these models have then been exported to biology. Fractals, reaction diffusion systems and diffusion limited aggregation are examples of physics being applied to biology. However, physicists seem to have been reluctant to look to the fundamental principles of biology and see how they apply to physics. The thinking seems to have been that Darwinian principles only apply to living systems - and that the high fidelity copying required is rare elsewhere. On this line of thinking, biologists have Darwinism covered.

In fact copying with variation and selection are ubiquitous in nature. Copying takes place whenever information at one location spreads to multiple locations. This can be formalized in terms of Shannon's concept of mutual information. The process is a common one. It happens when waves radiate. It happens when sunlight hits dust. Selection is similarly ubiquitous. In its simplest form, selection involves choosing a subset of items from a set. That happens all the time - and isn't confined to the death of living things or to choosing mates. When rocks erode, some of them collapse into the water, while others do not. Some of the Sun's photons hit planets, while others do not. These can usefully be interpreted as cases of natural selection.

Many simple physical systems behave as goal-directed fitness optimizers - in the same way as biological systems do. They can be used to solve optimization problems - much as genetic algorithms can. Examples of this are the way a lightning strike finds a short path to the ground or a stream in a mountainous region finds the fastest path to the sea. Physicists have noticed these optimization capabilities and have developed their own models of them - most notably maximum entropy production principle. These thermodynamic models and the Darwinian models cover similar territory and mutually illuminate each other.

Another area where Darwinian models have found to be applicable involves observation selection effects. Spencer's survival of the fittest can be usefully be generalized to observation of the observable - bringing observers into a central location within Darwinism. Classical Darwinian models of selection apply to observers in the exact same way that they apply to other entities.

One of the more studied areas of Darwinism within physics involves quantum physics. According to the many worlds interpretation, the world is constantly splitting. Some branches reproduce faster than other ones, and so become more numerous and thus more likely to be observed. Other interpretations propose a selection effect acting on these worlds - known as "wave function collapse". Whichever interpretation is closer to the truth, this looks as though this will be an interesting area for Darwinian or quasi-Darwinian models.

It has also been speculatively proposed that the visible universe is the product of Darwinian evolution - and thus has a lineage of ancestors which existed before the big bang. If so, we might reasonably expect to find some clues relating to this evolutionary heritage. We can see that the visible universe had a birth date - but other evidence of an evolutionary history currently remains elusive. We should do some further research on this possibility.

Darwinism is now over 150 years old. However, according to the picture here, the Darwinian revolution is only part way through. There's a lot of remaining revolution to go - and the process involves a lot more than just mopping up some creationists.

For references not hyperlinked to above, see the references here and here.